Credit card companies: how do they decide which of us are potentially profitable?
Most card companies use proprietary NPV models to determine potential profitability. These
models factor in revenue potential (propensity to revolve, merchant fees [some cards, like
'platinum' or 'elite' cards generate higher % fees than others] FX conversion fees, ATM / Cash
Advance fees, cross-sell potential or actual cross-sell revenue, etc); probability of write-offs
(which itself is a function of credit utilization rates, FICO, zip code, spending patterns, lifestyle
factors) due to non-payment; fraud w/o potential; service intensity (on line self service vs human
call center interactions; direct debit vs checks etc); cost of capital within the company; other
costs (e.g. the cost of keeping a $30K open to buy as opposed to $4K open to buy -- basically
opportunity cost of a % of that open to buy just sitting around in cash). Not defending the
companies, but it's all about profitability and these "black box" models can churn out some
unintuitive results. For example: a high spend, constant revolver at 50% credit utilization and
FICO 725 is probably much more profitable than a rock solid FICO 800, who is a moderate
spender but who refuses to revolve. The FICO 800 who simply keeps his or her high limit card in
the dresser drawer is a real loser (reserve expenses, chance of fraud, account maintenance, no
revenue). In good times, all these distinctions were invisible (cc marketers always thought they
could convert the non-users into users) but now, the companies have to staunch the red ink
Wall Street Journal Online Reader’s comment March 12, 2009
Comments by Keith Sharp:
“revolve” means not paying off full balance each month
‘FICO’ is a creditworthiness score
FX=foreign exchange (credit cards charge around 2%, quite a good rate for small amounts)
‘cost of capital’ used as ‘hurdle rate’ in PV (present value) calculations
‘open to buy’=unusued credit limit: issuer needs to keep a proportion of this liquid in case used